![[ Center for Humane Technology ]](https://substackcdn.com/image/fetch/$s_!uhgK!,w_120,h_120,c_fill,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5f9f5ef8-865a-4eb3-b23e-c8dfdc8401d2_518x518.png)
Last Friday, as I wrapped up my work week, I, like many of us, saw an unprecedented piece of AI news come across my newsfeed.
A government had taken the most capable AI models offline in a matter of hours.
The White House had sent an order to Anthropic, telling the AI company to suspend access to Fable 5 and Mythos 5 for any foreign national. The government cited national security concerns in its directive, and used export controls in order to get Anthropic to obey.
Details about the abrupt suspension are in flux, and so many variables have been at play — from Mythos’ eye-popping cyber capabilities, to the ongoing strained relationship between Anthropic and the White House. But what stood out to me more than anything were the implications of this event. Because in order to comply with the directive last Friday, Anthropic had to disable the Fable and Mythos for all users worldwide.
This event raises significant questions around who wields absolute decision-making authority around AI, and what that means for not just for AI companies, but for all of us. Might we face a future where the nationality on your passport is what gives you access to a technology — or restricts you from it? How comfortable should we be when one company or one government holds veto power over AI for tens (if not hundreds) of millions of people?
This latest development between Anthropic and the U.S. government tops off a remarkably busy — and turbulent — spring in the AI policy ecosystem. From the White House’s executive order on oversight of frontier models, to the ongoing push for liability legislation, policymakers are making good-faith efforts to shift AI outcomes in society. And while we appreciate these efforts to regulate this technology, they still have not resolved the fundamental uncertainty plaguing the AI ecosystem.
What we need is to move from a reactive regulatory approach to a thoughtful suite of regulatory tools — ones that provide clarity, consistency, and stability. One might argue this is difficult with a technology as volatile and fast-moving as AI. There’s truth to that, but as we note in CHT’s AI Roadmap, this suite of regulatory tools already exists — it just needs to be leveraged. And it includes government working with agencies and civil society to push forward safety and transparency standards for AI development — not as a one-off, but as a lever we can truly count on and build from.
I want to reiterate that a thoughtful approach to AI regulation doesn’t mean a slow approach. It means governance that allows technology to keep developing while ensuring the tech is actually developed well and deployed fairly. It means being proactive instead of scrambling to take entire products offline.
Events like last week’s with Fable and Mythos drive home the point that we need ongoing, robust conversations around what it will take to shift the incentives in AI development and deployment. Just as no one solution will be sufficient, no one perspective will be either. We need to keep expanding our discourse to make sure the analysis of AI is clear-eyed, deep, and able to spark lasting & incentive-shifting change.
That’s why I am excited to announce that CHT is launching its Emerging Voices in AI & Society Fellowship Program. This program will bring together experts with diverse professional and lived experience who view AI through an interdisciplinary lens. With focus areas including technologist perspectives, human cognition, relationships, surveillance, and spirituality, the Emerging Voices in AI & Society program will deepen the discourse around AI, and help us move from one-off reactions to incentive-shifting solutions.
Announcing The Emerging Voices in AI & Society Fellows Program
For AI to serve humanity’s best interests, we need more thoughtful, diverse voices shaping the story we tell about it. Today, we are opening applications for our new Emerging Voices in AI & Society Fellows Program to help carry that essential work forward.
If you feel you’d be a fit for our fellowship program, or know someone who might be, please visit our website to learn more. Applications open this week. We look forward to hearing from you and learning about your perspective when the next unprecedented AI news unfolds.
One last note: when people ask me how I stay levelheaded amid the deluge of AI events, I say simple human pleasures are what matter — so Happy World Cup to all fellow fans. I’ll keep bringing these conversations around AI to your inbox. And in the meantime, I’ll be screaming for my favorite teams out there on the pitch.
Julie Guirado
CHT Executive Director
 | A guest post by
|
Hi, I am long-term follower of CHT and go back on the philosophies even before the days of the Netflix special. I am a Lanier-head and follow many of his preachings.
Looked up the ideal candidate profile for the fellowship and noticed "Have an active on social media and are building a following around their ideas". My question is, CHT, what gives?
I think Bernie Sanders is onto something: AI should be managed as a common good where every citizen is a stakeholder. However, giving everyone direct ownership of corporate AI giants feels like a massive stretch (dividends aside!).
Knowing Bernie's style, it’s likely hyperbole meant to rattle tech executives and force them to look past the high-stakes race for AGI. If we want a realistic alternative to corporate "steamrolling," converting these entities into non-profit corporations—complete with strict safety guardrails and free public access to frontier models—would be a massive win for the public interest.